<aside> 🐙

Open on GitHub

https://github.com/nithinshettygit/AIRA-FINAL.git

</aside>

🚀 Problem Statement

In rural and under-resourced schools, there is often a shortage of qualified teachers and effective teaching materials. This gap leads to disengaged learning and limited access to quality education. AIRA (Artificially Intelligent Robotic Assistant) was designed to deliver interactive science lessons through an intelligent AI assistant that adapts to the student's learning style in real time.

🎯 Objective

To create an AI teaching agent that:

• Uses RAG pipeline and LLM(Large Language Model) to explain complex science topics
• Handles real-time student questions via chat interface
• Dynamically fetches YouTube videos and figures for deeper understanding
• Aligns with the Karnataka state science syllabus

🧰 Tools & Technologies Used

🧩 System Architecture

1. User Input →

Student enters a topic or question on the web interface via text/voice

2. Semantic Search + RAG →

Query is used to retrieve relevant information from a local knowledge base (using FAISS)

3. LLM (Groq LLM) →

Generates an explanation, summarizes, or answers questions based on retrieved chunks

4. Agent Control Layer →